subVI question

[commodore]

I am talking about a Q-bus DAC PCB here. The DEC itself works fine and does not give any noise that could interfeer with my analogue measurement.

So the noise is not from the DEC because I am not measuring that :).

[LV_Pro]

When you say you measure the pin or flat-cable, what are you measuring with? A DMM or other old style (i.e. Simpson 260 VOM meter) will show a different value on a noisy signal, they are performing averaging, whether intentionally or just through the physics of how they work. A better representation would be to look at the signals with a good,high speed analog, o'scope.  When you say, measure with respect to ground, well that has a lot of possible issues as well, particularly with low level signals.  There are a bunch of articles on noise in measuements on this (NI) site.

[commodore]

I am measuring with an DMM but that was not the problem.

In the meantime my supervisor had a look at it and he told me that the setup from the breadboard gave me noise.

The wires that I am using are not shielded whatsoever and are just hanging there sort of speak.  He told me not to look for the noise problem until the final PCB arrived.

I never worked with a PXI before and was not aware that it was so sensitive.

We don't have any engines or any other high demanding things at the lab HQ, only on site and that is far from here .

I will post it if my 20mV noise does not go away after I used the PCB instead of the breadboard. But my guess is that it will.

Thanks for the interest.

Stieve

[JÞB]

commodore,

with that many DAQmx tasks you do need to consider changing your approach to create each task only once.  After you have created 4096 tasks DAQmx MAY generate errors- I believe some background clean-up does take place for obsolete tasks (Those that use the exact same resources as previous tasks with auto-clean-up proprty = TRUE and are no longer in the commit state ) but it does take some processor time to create - validate, reserve, commit and start the tasks.   

You can optomize DAQ-mx task execution by creating and commiting the tasks initailly prior to entering the main loop. Like the model shown below

The DAQmx help file has some very good information that Sean N explained in detail in this Post.  Applying this information to an application with a large number of tasks will save a lot of hair-pulling and timing ussues.

[LV_Pro]

20 mV, in most situations, is not a surprising amount of noise. Noise comes from everywhere these days, with power, fluorescent lighting, etc. Where I'm working we had to turn off all cellphones, their signal, particularly as it is so high frequency, was being picked up by fairly short lengths of wire.

Connecting to make low noise analog signal measurements is non-trivial, as mentioned there are numerous discussions, white papers, etc., on National Instruments' site and elsewhere on the web.

Let us know if you have more problems, there is literally a world of experts on this forum, on a variety of subjects, with LabVIEW being the common thread.

[commodore]

@jeff.

You are right and I am working on that at the moment. My PC crashed because of that reason.

But the problem is that I only have until the 10th of June to get it running. So if the last version of my VI does not work, I will have to hand in the V1.0

Somebody already helped me on that earlier in this threat!

@LV_PRO

Indeed, noise is almost everywhere these days. But this amount of it should go away when I go from breadboard to PCB.

I am aware of the advantaged of this very useful forum. I honestly don't know any other company besides NI that has a forum like this.

And I will post on this threat if I still got noise on the PCB.

Bye,

Stieve

PS: I noticed the change of the topics name. That is probably for the best because this thread is becoming very useful for LabVIEW starters as myself!

[JÞB]

to point out specific areas of complex threads that the post addresses. And subject lines are filterable in searches

Tags also help find specific posts that contain tidbits you might want to visit again

[commodore]

Hi,

I got the system working on a interface PCB I made and the 20mV noise is gone.

It is reduced to 2mv - 5mV and my guess is that is acceptable.

Thanks for the support on the topic.

Greets,

Stieve